DESCRIPTION (Verbatim Adapted from Applicant's Abstract): Research has shown removing maximum amount of tumor with minimal sacrifice to normal tissues is the key to improving the survival rate of brain tumors. Thus, there is a greater need for an intra-operative tool, which effectively detects tumor margins in real time and provides a sub-millimeter spatial resolution for guidance of tumor resection. Optical spectroscopy can provide such a tool as it has the advantage of providing automated, real-time, non-intrusive diagnosis with high sensitivity and spatial resolution. Fluorescence and diffuse reflectance spectra were acquired from normal and various types of tumor brain tissues in vitro of about 20 patients. Based on the spectral differences, diagnostic algorithms developed showed that fluorescence could differentiate normal white and gray matter from primary tumors with 97 percent sensitivity. Fluorescence alone was insufficient in separating normal brain tissues from secondary tumors; combining diffuse reflectance with fluorescence yielded 97 percent sensitivity for this discrimination. Following the success of these studies, a pilot study of 21 patients was successfully performed. Preliminary in vivo results showed that tumor margin tissues can be differentiated from normal tissues with a sensitivity and specificity of 83 percent and 85 percent respectively using fluorescence and diffuse reflectance spectra. In this proposal, we plan to develop autofluorescence in combination with diffuse reflectance spectroscopy for intra-operative brain tumor and tumor margin detection in real-time to guide tumor resection. To achieve this goal, the following specific aims are proposed; (1) Characterize tissue fluorescence and diffuse reflectance signatures of brain tissues in vivo. (2) Develop diagnostic algorithms that separate normal and tumor tissues from tumor margins. (3) Study the basis of observed differences in the spectral characteristics using microspectroscopy, cyto-chemical analysis, and modeling. (4) Conduct retrospective and prospective evaluation of the algorithms developed to obtain estimates of their performance. (5) Assess the feasibility of using optical spectroscopy during stereotactic procedures and verify the performance capability of this technique for brain tumor demarcation. (6) Develop (a) software interface to implement and automate data acquisition and diagnosis that provides real-time feedback to the surgeon for therapy guidance and (b) next-generation clinical spectroscopic system to reduce the scale but not the accuracy of the spectroscopic system. This research will have tremendous impact on the future of tumor resection as upon the successful development of the proposed research, this can be translated to the application of other organ systems such as prostate and ovary.